With the sunsetting of Universal Analytics on 1st July 2023, marketers now have a deadline to make the move to Google Analytics 4. But, what’s different with GA4 and should you be concerned about how this will impact your reporting?
Google Analytics 4, initially ‘App + Web,’ was released two and a half years ago and is the latest iteration of Google Analytics software. In mid-2023, it will replace the current version, Universal Analytics.
In a previous blog, our Performance Director, Aaron, covered the changes in Google Analytics 4 at the time, however, the software is actively evolving so here are some more of the changes to be aware of.
Google Analytics 4 uses an event-based data model
Out with the old hits-based model, in with the new event-based model.
When it comes to Google Analytics 4, this is probably one of the more significant changes that many marketers might already be aware of. This model is intended for use across websites and apps with a user-centric mindset, which means better data on how users interact with your domain/app.
This means getting used to a whole new system and how we report insightful data. For technical users, this means no more notion of ‘hit types’; everything is the same, but with different parameters. In reality, this should be quite easy to adapt to because it is essentially Google Analytics catching up with how we work as marketers.
Yes, page views are a good metric, but do they add anything of value beyond this? To be user-centric, we must examine how users interact and engage with the pages of our website and GA4 will allow us to do this with new metrics, such as scroll depth, engaged sessions, and engagement rate.
As of writing, there are:
- Automatically collected events – basic interactions that are collected by default, i.e, ‘page_view’
- Enhanced measurement events – ‘opt-in’ events to collect additional data, i.e, ‘video_start’
- Recommended events – Events suggested by Google based on site context, i.e, Online sales – ‘add_to_wishlist’
- Custom events – Create or modify events using code or the Google Analytics configuration settings
Ideally, these would be set up with parameters that would be most important to you. For now, many of the metrics we are used to seeing in Universal Analytics are still here to stay within the event-based model.
Google Analytics 4 introduces data-driven attribution
If you’re already familiar with Universal Analytics, you’re probably familiar with the various cross-channel attribution models which are designed to give credit to different stages of your user’s journey to conversion. The new data-driven attribution model in GA4 uses machine learning algorithms to distribute credit to each of your marketing channels based on how users interact with different channels throughout their journey to conversion. This is designed to provide you with better insights and to help you understand what is performing the best for your business in lieu of accurate first-party data collected in the wake of the ongoing e-privacy crackdowns of the last few years.
As of writing, for attribution modelling in GA4, there is currently the:
- Data-driven attribution model
- Cross-channel rules-based model(s)
- Cross-channel first click
- Cross-channel linear
- Cross-channel position-based
- Cross-channel time decay
- Ads-preferred model
The data-driven model analyses the whole journey a customer takes from initial contact to conversion and assigns credit to varying events taken along the journey (i.e., social media ad > email > remarketing > conversion). The credit percentage is calculated based on account data. Overall, this model should better help you not only understand the value of your marketing efforts but also how your customers interact with events throughout the user journey.
Google suggests the data-driven model as the best option to improve your ROI, and they’re not wrong to suggest so. Since the data-driven model uses machine learning, it is the most user-friendly option and develops over time to provide more accurate results. Whilst this may seem like a win-win option, we highly recommend evaluating your user journey (and consulting your agency) to decide on the model that would best support you and your business objectives.
We recommend taking a look at all the options available and methodology for the range of attribution models featured in GA4 to best suit your needs.
Google Analytics 4 is built for a cross-platform landscape
Google Analytics has been around for just over 15 years and in that time, the search landscape has completely changed.
In 2020, 68.1% of all website visits came from mobile devices and in 2021, global consumer spending in mobile apps reached $133 billion, up nearly 20% from 2020. Universal Analytics was built to mainly report on data from a desktop perspective but the shift to mobile has caused the need to adapt. Launching GA4 as App + Web was definitely a key indication that Google was moving away from a desktop-only and single-property-only mindset toward a more user-centric data model.
Inevitably, the percentage of mobile-related searches and in-app purchases will continue to rise in the coming years. This is due not only to the increased accessibility of smartphones but also to the long-term impact of the COVID-19 pandemic. We’re making more decisions from our phones than ever before, and as marketers, we should be interested in how users interact with not only our websites but also our apps.
New dimensions and metrics available for reporting
If you were a keen reader and noticed the mention of scroll depth as a metric to track within GA4, you’ll be pleased to know that GA4 comes with the addition of new metrics and dimensions right out of the box.
Bounce rate reports have also been removed from reports by GA4. Why? While bounce rate is useful for understanding how many users leave your site after landing on your site’s first page, it can be difficult to attribute the cause. GA4 introduces a more useful alternative metric: engagement rate.
Google Analytics 4 anonymises IP addresses
GA4 will (eventually) no longer collect IP addresses. Currently, Universal Analytics collects IP addresses from users who visit their site with IP anonymisation or ‘masking’ being an opt-in feature for those outside of the EU.
With GA4, IP addresses are automatically anonymised and will eventually stop collecting IP addresses altogether. Whilst this may cause issues with some marketers, the need to collect IP addresses isn’t that important as GA4 has other tools available for marketers to take advantage of. Ultimately, this signifies Google’s commitment to building a privacy-first web.
Google Analytics 4 is not quite ready for full reporting
Google Analytics 4 is quite basic out of the box. However, with an entire year until UA is phased out and GA4 becomes the default, Google still has time to release updates to further develop it. It’s still important to migrate across to GA4 sooner rather than later and you can configure and customise it in advance to suit your needs, putting you ahead of the curve once everyone has been fully migrated over next year.
One key difference that we have highlighted is that it does not include all basic reports at the time of writing, such as landing page reports. We suggest that you make the most of the reports available to you in Universal Analytics whilst it still exists.
Universal Analytics data and Google Analytics 4 data will not look the same
Because of the technical differences between Universal Analytics and Google Analytics 4, data such as the number of sessions recorded in UA and GA4 may not match exactly.
This is another reason we recommend you use UA and GA4 in conjunction with each other before UA is sunsetted. This will not only save you some raised eyebrows when reporting GA4 data, but it will also provide you with the time to customise and configure your GA4 reporting to ensure you’re tracking what you really need.
If you would like an in-depth view of why data will not look the same, we recommend reading Aaron’s blog on everything new in Google Analytics 4. Or for further information, check out Google’s official support on this.
Whilst Google Analytics 4 will bring in some major changes to how we report data, in some ways, many things will be staying the same but we will have to become better at how we gather data. For instance, we will be moving towards a world that will have more emphasis on collecting 1st party data and we will be focusing more on user-centric metrics.
Should I upgrade to Google Analytics 4?
Yes, and if you want to measure YoY performance in 2023, then you should make the switch before 30th June 2022.